The Weekly Assessment of Child Behavior: Further Evidence of Reliability and Validity
Abstract
Background: A lack of a very brief, validated measure of externalizing problems has limited clinicians’ ongoing monitoring of treatment progress for children with externalizing problems. The Weekly Assessment of Child Behavior (WACB-P; Forte, Boys, & Timmer, 2012) was developed as a brief parent report to help monitor children’s externalizing problems in Parent-Child Interaction Therapy (PCIT). Little research, however, has examined the WACB-P’s reliability or validity. The purpose of the current study was to examine the WACB-P’s factor structure, internal consistency, test-retest reliability, convergent validity, and discriminant validity among children age 2 to 7 years, the target age group for PCIT.
Methods: Participants were 145 children from a pediatric clinic (a normative sample; 53% male; 41% African American, 21% Latino) and 36 children presenting at intake to a PCIT clinic for the treatment of externalizing problems (61% male; 56% African American, 33% Latino). Both clinics reside within a children’s hospital located in a low SES urban neighborhood in Philadelphia. Parents completed the 9-item WACB-P and the 36-item Eyberg Child Behavior Inventory (ECBI; Eyberg & Pincus, 1999), a well-validated measure of externalizing problems. The WACB-P uses a 7-point response format similar to that of the ECBI (sample item: “How often does your child do things right away when asked?”; 1= “never”; 7 = “always”) with the exception that higher scores indicate better behavior on the WACB-P but worse behavior on the ECBI.
Results: A principal components analysis suggested a one factor solution for the WACB-P that explained 46.4% of the variance. The WACB-P’s internal consistency was satisfactory as Cronbach’s alpha = .85. Test-retest reliability among a subsample of parents repeating the WACB-P in the pediatric clinic (n = 75; mean test-retest interval = 6 months) also was satisfactory (r = .64, p<.001). Evidence for convergent validity was found as the WACB-P was significantly correlated with the ECBI (r = -.60, p<.001). Evidence for discriminant validity also was found as the mean score for the WACB-P was lower (i.e., worse) for children presenting to the PCIT clinic than for those in the pediatric clinic (t(173) = 6.62, p<.0001; mean pediatric sample = 44.2, SD = 10.7; mean PCIT sample = 31.1, SD = 7.0).
Conclusion: The current findings suggest that the WACB-P is a promising measure with good psychometric properties based on its internal consistency, test-retest reliability, and discriminant validity. Only moderate support, however, was found for its convergent validity, raising the possibility that the WACB-P’s brevity and/or response format (e.g., assessing compliant behavior and emotional regulation vs. the ECBI’s assessment of non-compliance and tantrums) may have limited its strength of correlation with the ECBI. Future research is needed to examine whether there are specific dimensions of externalizing behavior (e.g., noncompliance; aggression) that the WACB-P best identifies among young children.